n the lesson on function overloading, you learned that you can create multiple functions of the same name that work differently depending on parameter type. Operator overloading allows the programmer to define how operators (such as +, -, ==, =, and !) should interact with various data types. Because operators in C++ are implemented as functions, operator overloading works very analogously to function overloading.
Consider the following example:
1
2
3
| int nX = 2;int nY = 3;cout << nX + nY << endl; |
C++ already knows how the plus operator (+) should be applied to integer operands — the compiler adds nX and nY together and returns the result. Now consider this case:
1
2
3
| Mystring String1 = "Hello, ";Mystring String2 = "World!";cout << String1 + String2 << endl; |
What would you expect to happen in this case? The intuitive expected result is that the string “Hello, World!” is printed on the screen. However, because Mystring is a user-defined class, C++ does not know what operator + should do. We need to tell it how the + operator should work with two objects of type Mystring. Once an operator has been overloaded, C++ will call the appropriate overloaded version of the operator based on parameter type. If you add two integers, the integer version of operator plus will be called. If you add two Mystrings, the Mystring version of operator plus will be called.
Almost any operator in C++ can be overloaded. The exceptions are: arithmetic if (?:), sizeof, scope (::), member selector (.), and member pointer selector (.*). You can overload the + operator to concatenate your user-defined string class, or add two Fraction class objects together. You can overload the << operator to make it easy to print your class to the screen (or a file). You can overload the equality operator (==) to compare two objects. This makes operator overloading one of the most useful features in C++ -- simply because it allows you to work with your classes in a more intuitive way.
Before we go into more details, there are a few things to keep in mind going forward.
First, at least one of the operands in any overloaded operator must be a user-defined type. This means you can not overload the plus operator to work with one integer and one double. However, you could overload the plus operator to work with an integer and a Mystring.
Second, you can only overload the operators that exist. You can not create new operators. For example, you could not create an operator ** to do exponents.
Third, all operators keep their current precedence and associativity, regardless of what they're used for. For example, the bitwise XOR operator (^) could be overloaded to do exponents, except it has the wrong precedence and associativity and there is no way to change this.
Within those confines, you will still find plenty of useful functionality to overload for your custom classes!
Operators as functions
When you see the expression
nX + nY, you can translate this in your head to operator+(nX, nY) (where operator+ is the name of the function). Similarly dX + dY becomes operator+(dX, dY). Even though both expressions call a function named operator+(), function overloading is used to resolve the function calls to different versions of the function based on parameter type(s). For example, In the lesson on arithmetic operators, you learned that C++ does integer and floating point division differently. This works because the operator/() function has two flavors — one that is called for integer operands, and one for floating point operands.
More generally, when evaluating an expression with operators, C++ looks at the operands around the operator to see what type they are. If all operands are built-in types, C++ calls a built-in routine. If any of the operands are user data types (eg. one of your classes), it looks to see whether the class has an overloaded operator function that it can call. If the compiler finds an overloaded operator whose parameters match the types of the operands, it calls that function. Otherwise, it produces a compiler error.
0 comments: